Open ghost opened 5 years ago
Hello everyone,
I am using the CycleGAN model on my horse2zebra dataset. I always get good results during training but when I am testing the model with some new images it always get distorted.
Is there anything that I am doing wrong ?
The training time performance is often better than the test time performance. One thing you can do is to add these new images as part of the training.
Hello everyone,
I am using the CycleGAN model on my horse2zebra dataset. I always get good results during training but when I am testing the model with some new images it always get distorted.
Is there anything that I am doing wrong ?